Contribution à l’étude des transferts dans des mortiers à base de kaolinite: de la perméabilité à la microstructure

Abstract

From an ecological and economic point of view, the present research aims to contribute to the valorization of calcined clays as supplementary cementitious materials (SCMs) in the composition of cementitious mortars. So, the objective of this study is to evaluate the influence of three SCMs which are: metakaolin (MK), calcined sediments (VC) and brick powder (PB) (added by substitution of cement with different dosages, ranging from 10 to 30%) on the physic-mechanical and durability properties of the tested mortars. In addition to a mechanical characterization of the mortars studied, a major interest is given to the evaluation of some transfer parameters in these modified mortars, such as open porosity, capillary absorption, water permeability, water vapor transmission permeability and carbonation (measured in accelerated and natural conditions). The impact of the carbonation of the material on the transfer properties was also investigated in this experimental study. The results obtained showed that the substitution of cement by 25% MK, 20% VC and 15% PB allowed obtaining the best mechanical strengths of the mortars. The MK-based mortar always showed the best performances, both in terms of physic-mechanics and durability, compared to the other mortars (based on VC and PB). Compared with the control mortar, the MK, VC and PB-based mortars ensured better durability performances. These three SCMs used decreased the porosity and permeability of the cementitious material, mainly because of their fineness and because of the Portlandite consumption by the pozzolanic reactions produced by the incorporation of these SCMs in the mortar. In contrast, this behavior can be contradictory in carbonated mortars: transfer properties in the mortar, such as porosity, can be higher in mortars based on MK, VC, or PB than in the reference mortar (without SCM). Consequently, the resistance to carbonation is also influenced in the opposite way. A phenomenon directly attributable to the microcracking effect influenced by shrinkage due to carbonation. The effect of carbonation on the transfer properties therefore depends on the initial mineralogical composition of the material. In order to simulate the carbonation depth of the different mortars studied, a model based on Artificial Neural Networks (ANN) was developed to predict the depth of carbonation of materials based on MK, VC and PB. The prediction model was developed using the experimental results of carbonation measured in accelerated and natural conditions. The model's development took into account parameters related to the characteristics of the mixture (cement content, SCMs content, mortar age) and parameters related to environmental conditions (CO2 concentration, relative humidity, temperature and exposure time to CO2). The learning, validation and testing performance of the developed ANN model showed very good correlations, exceeding 90%, between the experimental values and the values predicted by the model.